Distributed Zonotopic Fusion Estimation for Interconnected Systems With Communication Constraints

This paper investigates the distributed zonotopic fusion estimation problem for large-scale interconnected systems subject to unknown-but-bounded noise and communication constraints. To cope with limited resources in sensor-to-local-remote estimator (S-LRE) channels, an event-triggered coding–decoding mechanism under bit rate constraints is introduced to reduce communication load. To exploit overlapping state redundancy induced by system decomposition, a zonotopic fusion criterion is proposed to yield fused estimates that are superior to local estimation. With the zonotopic interval width selected as the performance index, the estimator gain matrix and the fusion parameter matrix can be efficiently computed via parallel linear programming problems, and the optimal bit rate allocation for each S-LRE channel is obtained through offline optimization. Moreover, robust stability conditions are derived to guarantee the ultimate boundedness of the maximum interval width of the local zonotopic generator matrices. Finally, a system of four interconnected mass–spring–damper vehicles is employed to validate the effectiveness and superiority of the proposed methods. © 2026 IEEE.